shebs@utah-cs.UUCP (Stanley Shebs) (10/12/83)
So what *specific* actions can we take to head 'em off at the pass? I suppose we could start a letter-writing campaign to Congress, but on the other hand, do you want to publicly reveal your use of a modem before the issue is resolved? Interested in ideas, stan the l.h. utah-cs!shebs
lai@allegra.UUCP (10/13/83)
The key question is whether it costs the phone company more to provide service to people who have installed modems. I have been using a modem from home for many years and I know the answer is yes - I often use my modem for hours on end, locking up a line in the central office, and yet I pay the same low flat rate as a residential customer who probably uses the phone only 10% as much I do. It is an unfair subsidy as far as non-modem phone users are concerned. I heard that some phone companies are starting to address this situation by timing local calls and charge customers for total connection time above a certain limit per month. This is probably a fairer solution than to charge all modem users the flat business rate since not all modem users use their lines with the same frequency.
warren@ihnss.UUCP (10/13/83)
(Views my own, not those of my employer) First, as of January 1, 1984, there is no more Ma Bell. The company you will be dealing with regarding this is your local phone company, which, by the way, is not the part of the old AT&T that was de-regulated to enter new ares. (Though I believe that the new regional phone companies do have more leeway than before). The question of what rate people should pay is not a simple issue. The two principles that have guided this have been pay in proportion to value received and in proportion to the cost of service. As a modem user, you are receiving a different service, and thus different value. It is also quite likely, as someone pointed out, that you are using considerably more resources than someone with just a plain voice line. If you had measured service, this difference would be reflected in your bill, but if you have flat rate service, it won't. You can and should consider writing your local public utility commission about whether or not the particular charging scheme being used is fair, however it is clear that a someone using a modem on a local phone line several hours a day should get a different monthly bill than someone using the phone only to place one or two calls. The fair charge may ultmiately be even more than the business rate! There is another effect of artifically low rates for certain services that data communication users should consider before complaining. Modems are a particularly inefficient way to carry data in a modern telephone network, which internally is frequently using 64,000 bits/second to carry your call, even though you get only 300 or maybe 1200 bits/second out of it. There are many emerging technologies for making higher speeds available, however services based on these technologies are priced closer to their true cost. If data communication using modems is available below cost, then the more efficient techniques may not be cost competitive, and manufactures will not get the volume to make low cost, high speed data communication available. I am sure that everyone out there in netland would like faster, cheaper communication. Once again, these are my personal views, not necessarily those of AT&T Bell Laboratories. -- Warren Montgomery ihnss!warren IH x2494
hal@cornell.UUCP (Hal Perkins) (10/14/83)
Charging modem users for connect time isn't all that great either.
When I had a terminal at home, a lot of the connect time was idle
time--just like any other situation where I use a terminal. Much
of the connect time is spent reading and thinking, not transmitting
data one way or another. I'd hate to have to keep hanging up the
phone, then redialing and logging in again to avoid connect charges.
I would have no objection to paying on the basis of the amount of
data transmitted. But I do object to paying for connect time or
paying business rates. It is true that modem users tie up voice-
grade lines for long periods of time, but that's no excuse. My
impression is that the local loop to my central office is used for
nothing else, so it shouldn't matter if I use it for 10 minutes or
10 hours. And the phone company ought to be able to use some sort of
multiplexing techniques to overlap my "think time" with someone else's
data transmission time--whether the other data comes from a digital
device or a digitized voice conversation.
Allowing the phone company to charge for connect time simply rewards
them for using old technology that requires tying up physical circuits
for the entire duration of the terminal session. The equitable solution
is to tell them they can charage based on the amount of information
transferred, and leave it to them to figure out how to make most efficient
use of their inter-office wires.
Hal Perkins UUCP: {decvax|vax135|...}!cornell!hal
Cornell Computer Science ARPA: hal@cornell BITNET: hal@crnlcseric@aplvax.UUCP (10/14/83)
You do have a good point about phone usage, but on the other hand, we all subsidize the equipment needed to handle the Mother's day rush, which I personally do not take advantage of. I am not saying this is bad, just that you can't get rid of all inequities. BTW, any estimates on just how many people do use modems from home? I would imagine it is still a very small percentage compared to number of residential phone lines. -- eric ...!seismo!umcp-cs!aplvax!eric
leichter@yale-com.UUCP (Jerry Leichter) (10/14/83)
I don't like the idea of Bell charging a lot for home modems, but I dislike uninformed flames like this one even more. Hal Perkins is uninformed both about technology and economics: Technology: It is quite true that the local loop between your home and your local office is used by you only; however, there is obviously other equipment involved in making your call. Something has to connect your local loop to the other end of your conversation. What that "something" is depends on the type of equipment in your central office, but there are always limits on central office interconnects that are a lot smaller than "every circuit active". The limits are there to keep the price of the equipment reasonable. It may be that the costs will change, and that new technologies may even make it possible to just allow every line to be active, but that's not the case today and is not likely to be soon. (As a simple example of resources you use whether you talk or not: Where do you think the power to drive your phone comes from?) Economic: "The ought to be forced to put in a technology that will allow charging by bits sent, etc." What an amazing statement! You, a member of a small minority of people who use the system in an idiosyncratic way, want to impose the enormous cost of replacing billions of dollars of capitol equipment, which is otherwise good for another 20 years or more, on everyone else. I remember bitches from years gone by about "Long distance would be much cheaper if Bell would use sattelites but they refuse because they have all this money invested in microwaves". Yes - so what? The costs of setting up a national network are IMMENSE, and can only be justified by assuming the equipment will last for quite a while. That means you can't throw everything you have out for the latest do-hicky. These new features are becoming availa- ble now because competitors like MCI and Tymnet are moving in with no existing capitol investment and applying new technologies to exactly the customers who have a use for it. (I know this sounds like AT&T's "cream-skimming" argument against competitors; it is. As a general rule, I dislike monopolies and regulation, but I must say that in the case of AT&T the likely course for telephone service in the near future is toward poorer, more expensive service, on average, as a result of the breakup.) If you want to pay for connections on the basis of bits sent, get a Tymnet or Telenet account. Of course, you still have to get to them by phone...sorry, you lose, until someone puts up the money to pay for the new equipment to provide you with another option. TANSTAFL - someone has to pay for that stuff - why do you think it should be anyone but you? BTW, I have never worked for AT&T or any of its constituent parts. -- Jerry decvax!yale-comix!leichter leichter@yale
nazgul@apollo.UUCP (Kee Hinckley) (10/14/83)
> on the other hand, do you want to publicly reveal your use of a > modem before the issue is resolved? It's my understanding that Ma Bell has the ability to detect modem usage on the phone lines. Which means there is really no way to prevent them from knowing about your modem. Verification? -kee
sef@druxu.UUCP (Farleigh) (10/14/83)
Hal Perkins made two comments about Cental Offices that are
incorrect.
1. The Telco's can't "multiplex" data from customers (this
technique is called "packet switching") by order of the
Fed's. I do not know if this restriction will still be
valid after the break-up of the Bell System.
2. While that pair of wires coming into your house from the
CO is for your use only the CO cannot handle every customer
being off hook, on the telephone, at the same time.
This is known as a blocking switch.
Scott E. Farleigh
AT&TIS Labs Denverkurt@fluke.UUCP (Kurt Guntheroth) (10/14/83)
Bell wants to charge more for modems because it costs them more to service
calls that use modems. Here's why.
A typical call between two humans lasts only a few minutes. Calls between
modems can last for hours. The phone company has less trunk lines than
telephones because not everybody is making a call at the same time. More
long 'conversations' between modems means the trunk lines are loaded more
heavily, thus invalidating the assumptions on which the rate structure is
based, and forcing the phone company to add capacity.
Some calls (long distance) are multiplexed over the communication channel.
It is very typical that one person is speaking and the other person is
listening. It is not necessary for the dead air of the listening person to
be transmitted. Instead, the listener's half of the channel is allocated to
another call and quickly restored when the listener begins to talk (You can
occaisionally notice a situation in which the first syllable of someone's
conversation is cut off when you call long distance. This is the phone
company not switching channels fast enough). Modems, on the other hand,
both 'speak' continuously, forcing both parts of the communication channel
to be kept open, which again increases the cost over what is expected of a
'normal' call.
So it is not surprising why the phone company wants to charge you more if
you use a modem. This is not to say that it is good public policy to let
them do this. Maybe if they only charge extra for individual calls that use
a modem... Certainly they should not be allowed to have a monopoly on data
communication.
Kurt Guntheroth
John Fluke Mfg. Co., Inc.
--
Kurt Guntheroth
John Fluke Mfg. Co., Inc.
{uw-beaver,decvax!microsof,ucbvax!lbl-csam,allegra,ssc-vax}!fluke!kurtsmb@ulysses.UUCP (10/14/83)
Modems are essentially parasites on the voice phone network. That is, the network was designed, engineered, and intended for adequate-quality voice communication. Modem design is the art of transforming bits into whistles, squeaks, and buzzes that have characteristics compatible with the existing network. Given that, it also makes sense for the phone company to build bandwidth-saving devices that operate on the assumption that speech is what will be carried. Now -- according to some statements from AT&T that I saw published long before I joined this august establishment, about 85% of the local loops are physically capable of carrying 56Kb/sec data -- *if* the terminating equipment is intended for that purpose. I suspect that they'll try to make such services available. But new services tend to be priced at somewhat closer to cost than POTS (plain old telephone service) is, which means that it won't be competitive unless modem users are paying their share. I should point out, of course, that "think time" is not idle time for the phone line with a conventional modem; it's still sending out its normal carrier. (Incidentally, local loops will not necessarily be dedicated forever. Technology is at, or close to, the point where multiplexing on a trunk to a neighborhood makes sense. Modem can screw that up.) I'm no fonder than anyone of paying more for phone service. I had a separate phone line for my terminal well before I joined Bell Labs, so I know what it would have been like to pay more on a graduate student budget. And the tarriff in question, in Oklahoma, was filed in 1965, when it was reasonable to make different assumptions about usage patterns. But it doesn't seem at all unreasonable to me to levy some surcharge for modem use. --Steve Bellovin Disclaimer: these opinions are mine, and in no way reflect any policies or non-policies of ATT, ATT Bell Laboratories, NJ Bell, the FCC, or any other sentient entities.
dee@cca.UUCP (Donald Eastlake) (10/14/83)
I don't understand what the big deal is on this. Anyone who
wants to know what the official rate structure of the phone
company is can read the traffis that are normally required by law
to be on file in every phone company business office. The
present structure tends to charge more for business service and
more for data service. It always has. How the rates are skewed
is more or less a political decision which, up to now has always
favored the small/poor non-data residence user. If enough
political pressure were applied, perhaps the rates could get
skewed to favor data user who tend to have very much longer
holding times and want cleaner lines that the residence users.
If you want to aply such pressure, to your PUC and/or
legislature, feel free to do so. But I don't see why you make
out the present situation as some big conspiracy by the Phone
Company.
+ Donald E. Eastlake, III
ARPA: dee@CCA-UNIX usenet: {decvax,linus}!cca!deehlh@linus.UUCP (Henry L. Hall) (10/15/83)
Still another possibilty to "get out of this entanglement" would
be to continue the current rate structure for modem users during low usage
hours. If you wanted to use your modem during peak hours then you would
have to pay for it. During off-peak hours, however, I don't see the
justification for higher prices except to keep the billing algorithm
simpler. Lines used for peak hours could then be utilised more fully
during off-peak hours without requiring the installation of more lines
or a different routing pattern. After all, since most connections are
routed automatically, (how much did your phone service actually degrade
during the telephone workers strike anyway, other than queues for the
operator and directory assistance) the only costs involved for Bell with
the use of modems should be the energy used to carry the signal, and didn't
Bell figure out what that cost was going to be to handle the current number
of telephone lines.
Trying to be fair (But still a midnight hacker)
And I even pay my phone bill on time,
Henry Hall
{allegra, cbosgd, decvax, ihnp4} !linus!hlhkarn@eagle.UUCP (Phil Karn) (10/15/83)
Local data services and the alternatives is one of my favorite topics; I'd like to add a few comments of my own to the fray. Clearly, charging for having modems PER SE connected to your phone line is unfair; I suspect that this particular tariff will get thrown out not only because it is unreasonable but also because there is no way for the telco to detect them without monitoring phone calls. True, you're supposed to report direct connect modems, but not accoustic couplers. What is most likely to happen in Oklahoma if people complain enough is that the "modem charge" will get replaced with steep connect time charges. In other words, the computer users will have a relatively easily evadable charge replaced with an unavoidable one, probably just as large if you use your modem a lot. Clearly, the present situation is a bad one. The telcos have a large amount of plant invested that is designed for voice traffic and there is also a growing need for local data service which is uneconomical to handle via the old network. There are two general ways out: innovation by the Telcos and bypass by non-Telco carriers. In the first category is Local Area Data Transport (LADT) which is being developed for a home information field trial in Florida. This uses supersonic FSK modems on your local loop to carry full duplex data at 4.8 kbps to a packet switch at the central office. I do not know how if or when this service will become generally available. Clearly, however, if the Telcos are going to provide an alternative to dialup modems, this will be it. In the second category there are several alternatives: two-way cable TV and packet radio. While all the technology has existed for the first (Sytek, etc) for quite some time, I don't know of any regular CATV company that is providing this kind of service on a general basis. (If anybody knows of one, I'd very much like to know the details.) I would speculate that the average CATV company is more interested in selling pay TV channels than in experimenting with an exotic, special interest technology for which they don't know how to charge. Selling two-way services would also move them a step closer to being considered common carriers, something they've been trying to avoid for some time. Packet radio has the interesting characteristic of being something that you can take the initiative on instead of waiting for somebody else to come to your door selling a data service. If you have the equipment and can get an FCC license, that's all you need. While there is no "Personal Packet Radio Service" yet, I believe the FCC is now in the most permissive mood it has ever been toward innovative new services. If the system is strictly experimental and is not to be used for business, then all you need is an amateur radio license. The FCC greatly liberalized the rules for digital data transmissions in the amateur service a while ago, so that they are no longer a limiting factor. Quite a bit of original packet radio work is already taking place in the amateur service, and I think you will see some impressive things coming out of it. If you want to experiment with packet radio for business, it should not take too much effort to get an experimental FCC license if you do your homework and can prove it to the FCC. Of course, the fact that the license is "experimental" means that the technology isn't really all here yet, so if you just want simple, hassle free data communications, then you're still probably best off sticking with the telephone! The opinions expressed herein are strictly my own. Phil Karn
jlw@ariel.UUCP (J.WOOD) (10/15/83)
I think that two-way CATV is, at present, a highly overrated possibility for local data communications or TELCO bypass. Since 50% of all homes now have cable out in the street, the CATV companies are not the brash young upstarts everyone credits them with being. They have one heck of a lot of plant investment. This is an investment in plant which just wasn't engineered for two way for one thing and for another the AGC or level requirements for RF data units are orders of magnitude more stringent than for TV sets. You just can't go 20dB and then add an amplifier. They are in the same heavy investment category as the TELCOs are. Joseph L. Wood, III (201) 834-3759 ariel!jlw
cjh@ihuxr.UUCP (C. J. Holzwarth) (10/16/83)
The question of how the phone co will know that you are using a modem is a good one. They can monitor your line and find out. If they should do this though they are breaking several federal laws. Monitoring the information being transmitted on a telephone line is also known as wire tapping. There are also laws reguarding invasion of privacy. The only legal way of obtaining information about modem use is when someone complains that their modem doesn't work of requesting a court order for a wire tap of the population of the city. As far as paying for connect time, I'll pay for my connect time if everyone else has to, but if I don't want to pay for a data line and am willing to live with data transmission over a voice line, I shouldn't have to pay for the data line. 1984 isn't fat off, C. J. Holzwarth ihuxr!cjh
henry@utzoo.UUCP (Henry Spencer) (10/17/83)
The installed plant of cable TV outfits is also, in many cases, in
amazingly bad shape. I saw a paper recently on some of the problems
involved in using existing CATV hardware for data communications; the
biggest single hassle is that CATV outfits have immense networks of
relatively sleazy hardware which is poorly maintained and often operating
in adverse conditions. It's mildly amazing, actually, that they usually
manage to get a decent TV picture through. Not a good bet for data use.
--
Henry Spencer @ U of Toronto Zoology
{allegra,ihnp4,linus,decvax}!utzoo!henrykurt@fluke.UUCP (Kurt Guntheroth) (10/17/83)
[ Why can't the phone company reuse my 'think-time' for other calls and not
charge me for it? ]
Your modem sends a carrier out whether you are sending characters or not.
So does the modem at the other end. These signals must be present. If a
modem stops sending, the modem at the other end is apt to hang up.
So you see your phone is active the whole time whether you are transmitting
usful stuff on it or not. The phone company has a right to charge you for
this resource since it cannot reclaim it.
Maybe we need a different kind of modem?
--
Kurt Guntheroth
John Fluke Mfg. Co., Inc.
{uw-beaver,decvax!microsof,ucbvax!lbl-csam,allegra,ssc-vax}!fluke!kurtbarmar@mit-eddie.UUCP (Barry Margolin) (10/18/83)
Since one of the big problems that modems cause for the current phone system is that it wasn't designed for people who make long calls, perhaps a reasonable rate structure would be one which charged more per minute after the first 20 or 30 minutes. This way, modem owners would get normal rates when they make normal phone calls. Of course, this is incompatible with flat rate service, which is currently offered in most areas. However, I have heard many reports that the operating companies are planning on getting rid of this, anyway. I really doubt that many people actually make normal calls on the phone that their modem is on. Most people like to be able to get and make normal calls while they are dialed up, so they get a second line for their modem. (I currently have such a setpu, because my company hasn't gotten around to installing a line for me, and I find it quite inconvenient.) -- Barry Margolin ARPA: barmar@MIT-Multics UUCP: ..!genrad!mit-eddie!barmar
mason@utcsrgv.UUCP (Dave Mason) (10/18/83)
I wouldn't mind paying a modem charge if it meant I got real digital
transmission. The bandwidth utilization of even a 4800 baud digital
line is 8% of a digitized conversation line, and I would love to pay
8% of my current bill for 4 times the speed and clean lines. IN FACT,
I'd even be willing to pay a little more (to amortize the Telco's cost
in physical plant) than the current rate.
-- Dave Mason, U. Toronto CSRG,
{cornell,watmath,ihnp4,floyd,allegra,utzoo,uw-beaver}!utcsrgv!mason
or {decvax,linus,research}!utzoo!utcsrgv!mason (UUCP)leichter@yale-com.UUCP (Jerry Leichter) (10/18/83)
On the question of how Telco can tell if you are using a modem: Telco's have always had the right to listen in to check for equipment faults, monitor line quality, even check for fraud if need be. The general rule is that your LINE isn't private, even though your CONVERSATION typically is. For example, it used to be that if you called the operator and com- plained that a number had been busy for a long time, the operator could check by just tapping in to the line and listening. More modern equipment provides an automated tap - but through a filter that makes it impossible to understand the speech overheard. It's hard to see how this could be considered intrusive by a court or even a regulatory board. The fact that you are using the line - the only thing that can be determined this way - is knowledge the phone company obviously already has. To determine if you are using a modem, Telco need only connect a device that is sensitive to the modem carrier frequencies to your line for a very short period of time. While you may not like the result - a higher phone bill - I claim it is rather hard to argue that your privacy has been compromised. Remember, what you use for phone for has ALWAYS been a factor in billing - businesses pay MUCH more. -- Jerry decvax!yale-comix!leichter leichter@yale
stevesu@bronze.UUCP (Steve Summit) (10/19/83)
Although I hope they don't choose to do so, the phone company can
quite legally determine if you're using a modem. In fact, they
do it already, and you should be glad they do. Long distance
trunks all have devices called echo suppressors. These disable
transmission in one direction when something is going in the
other. Without them, your voice would bounce off the other end
and you'd hear it a second later. (Contrary to submissions
elsewhere on the net, there can be a lot of propagation delay on
a phone call, particularly if it travels via satellite.
Furthermore, older local trunks are often heavily loaded, which
improves the audio characteristics at vocal frequencies, at the
expense of a significant delay and severe bandwidth limitation.
Modems don't work at all on loaded lines) When you realize that
phone calls are bidirectional and have LOTS of gain, it's
surprising that there isn't more echo and feedback. It's due to
echo suppressors that it's usually hard to interrupt someone on a
long-distance call. They simply can't hear you.
Anyway, the point of all this is that an echo suppressor would
obviously keep two modems from talking to each other, since each
has to continuously send the other a carrier. For this very
reason, each and every echo suppressor in the entire phone system
has a modem detector in it, which turns off the echo suppression
in the presence of sustained carrier frequencies.
When you stop and think about it, the phone company monitors your
call in a lot of other ways, too. They notice when you hang up,
so they can release the common circuitry. They keep track of how
long the call lasts, so they can bill you. I've heard that some
central offices notice if you're using a touch-tone phone (since
the vast majority of lines now seem to support them) and ask you
to pay for touch-tone service if you're not already. None of
this is 1984.
It certainly isn't fair to unilaterally charge modem users more
for their local connect time. I know I spend far less time on
the phone using a modem than my little brother does talking to
his girlfriend. The argument that a long-distance modem call
ties up two trunks all the time while conversation can be
multiplexed is a good one. However, the proposal being discussed
is for the local office to charge more for connect time for a
local user, which (after January 1) has nothing to do with the
long-distance connections.
Audio modems are in fact incredibly wasteful, especially on
long-distance trunks. After your modem cleverly disguises 300 or
1200 bits/second to look like an audio signal (yes, I know baud
!= bits/sec, but it's close) the phone company turns right around
and digitizes the audio, at a MUCH higher baud rate. Even when
direct data connections become available (and this will certainly
happen soon) I'll still need my audio modem becqause I call up a
lot of rinky-dink computers that will keep their audio dialups
for a long time before converting. At risk of repeating what's
been said already (I hate people who do that!) it would be fine
to charge people more for using more connect time, but not just
for the fact that their audio is generated by a modem and not
their vocal cords.
Steve Summit
tektronix!tekmdp!bronze!stevesurf@wu1.UUCP (10/19/83)
Further comments on this matter can be found in net.dcom, please look there. Randolph Fritz Western Union Telegraph
condict@csd1.UUCP (Michael Condict) (10/20/83)
By my reckoning, a 4800 baud modem link requires a bandwidth of at least 9600 Hz (assuming one avoids any inefficiencies associated with translating between digital and audio), while the phone co. currently only guarantees its voice lines a frequency response up to approx. 5K Hz, which is certainly less than that. Thus, where do you get this figure that a 4800 baud modem only needs 8% of a digitized voice line? And why does it matter whether the line is digitized or not? You can't beat information theoretic restrictions. To get a frequency response up to 10K Hz you need a band-width of at least 10K Hz, regardless of how you code the information (unless of course you take advantage of some knowledge of the signal, such as the redundancy of human speech, to reduce the quantity of information that needs to be trans- mitted -- but that wouldn't help you in this case, because the redundancy in human speech compared to modem transmissions means that even less bandwidth is required for the voice line as compared to the modem). If I have displayed a grievous lack of knowledge of communications here, please just tell me by mail, to avoid embarassing me and boring others. If not, please tell me what's wrong with my analysis (which, by the way, indicates that 2400 baud modems are on the theoretical limit of error-free transmission with a 5K Hz bandwidth and should therefore not be expected to be very reliable; and 4800 baud modems should require two independent phone lines, which I have heard that they do). Michael Condict ...!cmcl2!csd1!condict N.Y.U.
fred@umcp-cs.UUCP (10/25/83)
From: stevesu@bronze . . . Long distance trunks all have devices called echo suppressors. These disable transmission in one direction when something is going in the other. Without them, your voice would bounce off the other end and you'd hear it a second later. . . . Steve Summit And if you don't beleive this, you should try talking on a long-distance line on which the echo-supressors are malfunctioning! I had this happen once, and found it very disconcerting. Sort of like having a short-circuit in the brain. It becomes totally impossible to hold a conversation. I'd like to have something to intentionally cause this sort of echo, so I can switch it on when I get a call from a particularly obnoxious phone salesperson. Fred Blonder harpo!seismo!umcp-cs!fred